Music provides an ideal domain to test the generality of widely accepted principles of cognitive aging, and to explore how the factor of expertise can moderate age differences in cognition. We propose 12 studies that will examine how age and musical training are related to perception and memory of music. Our first specific aim is to examine the hypothesis that one's knowledge of scales and other musical structures does not decline and may even increase in old age, despite age-related deficits in on-line analysis of music. We will test this hypothesis in five different studies that assess the effects of tonal structure on (a) the perceived similarity of single tones and tone sequences, and (b) accuracy of short-term memory for melodies, by young and elderly adults with different levels of musical training. Our second specific aim is to determine if the principle of generalized slowing in old age extends to musical materials, and, if so, if this slowing is reduced among expert musicians. Avoiding the use of response-time measures, our approach will be to determine the fastest and slowest tempos at which melodies can be recognized, and to assess the accuracy of pitch-encoding in melodies presented at varying tempos. Our third specific aim is to examine age differences in several aspects of memory for melodies, including the maintenance of melodies in short-term or "working" memory, recognition of melodies, recollection of context of prior encounter with melodies, and reality-monitoring judgments of whether familiar melodies have been recently perceived or simply imagined. We will also test the principle that there are minimal age differences in "priming" tasks revealing memory for stimuli even in the absence of conscious recollection. Of interest throughout the proposed research will be possible interactions of age with expertise, and, in other cases, dissociative effects of these variables.